Idea/Request: statistical power as a telescope

[DanChaltiel]

Hi Allison,

Thank you for your incredible pictures, they are a real inspiration and a wonderful teaching material.

I don't know if you accept such requests (I would totally understand if you don't), but I have an idea of drawing for base Statistical knowledge (somehow like your type I and II error drawings).

I found that a number of students had trouble understanding the concept of statistical power. I usually explain it to them by comparing the power of a statistical test to the power of a telescope:

• There are lots of stars in space, more or less large (near or far).
• The more powerful the telescope is, the more chance you have of seeing a small star.
• If you cannot see anything, either there are no star where you are looking, or your telescope is not powerful enough. You never can draw a 100% certain conclusion about this.

It seems that this comparison helped them a bit. I would love to see it as an image, but my tremendous lack of drawing skill prevents me from doing it myself.

For instance, I could imagine a scene where "data" is looking through a telescope at the center of a "relationships" star cloud where there is nothing to be seen, with the caption "maybe I'm not powerful enough" or something similar. I'm not sure how to represent those characters though, I'd let that to the artist 😄

Have a nice day
Dan